Looking for beta testers of new Polar Alignment utility.

[themos]

This was from the weekend, 450D with 200mm lens, S resolution, about 10arcsec/pixel, showing the development version.

And this is what PHDLab made of the subsequent session (note the Dithering spikes). PHDLab independently computed the polar alignment error at 0.61 arcminutes, matching PPA's estimate (to exceptional precision!).

Edited December 15, 2014 by themos

[StevieDvd]

Hi Themos,

Sorry for not getting back quicker on the v1.03 error O enocuntered. I started ppa.py in Anaconda console as suggested and got this:

The system cannot find the path specified.

C:\Users\Steve\Anaconda>cd C:\Users\Steve\Documents\PhotoPolarAlign

C:\Users\Steve\Documents\PhotoPolarAlign>ppa.py
Traceback (most recent call last):
  File "C:\Users\Steve\Documents\PhotoPolarAlign\PPA.py", line 1501, in <module>
    APP = PhotoPolarAlign(master=ROOT)
  File "C:\Users\Steve\Documents\PhotoPolarAlign\PPA.py", line 1495, in __init__
    self.Sniff_AstroTortilla()
  File "C:\Users\Steve\Documents\PhotoPolarAlign\PPA.py", line 1305, in Sniff_AstroTortilla
    lfn = os.listdir(dir)
WindowsError: [Error 3] The system cannot find the path specified: 'C:\\Users\\Steve\\AppData\\Local\\astrotortilla.sf.net\\AstroTortilla/*.*'

Not having dabbled in Python before I can't see if there is a new config setting I'm missing to enable/disable local solving.

I'm using Win 7 pro as OS and don't have Astrotortilla installed - but do have Astroart5

Hope that's a genuine bug I found for you and not me doing something stupid!

[themos]

Thanks Stevie, every little helps. If you can, open PPA.py in an editor  and go to line 1305 and change it to 

           #  lfn = os.listdir(dir)           lfn = []

keeping the same indentation as before.  That should allow you to carry on with the program.  It will be interesting to see if AstroArt can work with PPA. You will need to save images as JPEG format and also save the plate-solved .fits file that AstroArt produces and rename it with a .wcs ending (so there would be, say, img2014.jpg and img2014.wcs files)

Edited December 15, 2014 by themos

[StevieDvd]

Thanks Stevie, every little helps. If you can, open PPA.py in an editor  and go to line 1305 and change it to 

           #  lfn = os.listdir(dir)           lfn = []

keeping the same indentation as before.  That should allow you to carry on with the program.  It will be interesting to see if AstroArt can work with PPA. You will need to save images as JPEG format and also save the plate-solved .fits file that AstroArt produces and rename it with a .wcs ending (so there would be, say, img2014.jpg and img2014.wcs files)

Thanks, I worked around the missing directory once I saw the cause.

BTW, I also meant to mention that when using a mono camera the overlays are also being saved in mono. As a quick test I added  .convert("RGB") to the image open (s) and got the overlays in colour. If that's a safe thing to do without checking the incoming file colours.

[themos]

Thanks, that is useful. Let me know how you get on with AstroArt.

[StevieDvd]

Managed to spend a bit of time getting the alignment as good as I could before the hour got too late or the clouds started to interfere.

I hope this is accurate enough to use when I get my guiding setup finished.

Once you get the feel for the amount and direction of changes to make on the mount alt/azz bolts it becomes quite easy. I was using a DMK21 with focal reducer and live view so I could almost work out the rough position the stars needed to move to beofre taking and uploading an image. Would have been a lot easier if my computer was nearer the scope, the latop was hooked up in the garage so was popping in/out each time to check improvements.

Thanks Themos that's a real nice utility for alignment

Steve

[themos]

Thanks Stevie, that looks like a tolerable error, I am not sure I could do better. I am right-handed, so I stand east of the mount, looking north, and to turn the polar axis right, I need to turn the east azimuth knob clockwise (having loosened the west one). I should probably measure how much angle a half turn buys me. I expect each person will make their own mental notes.

Themos

[StevieDvd]

Thanks Stevie, that looks like a tolerable error, I am not sure I could do better. I am right-handed, so I stand east of the mount, looking north, and to turn the polar axis right, I need to turn the east azimuth knob clockwise (having loosened the west one). I should probably measure how much angle a half turn buys me. I expect each person will make their own mental notes.

Themos

I had a slightly larger error, but small enough I thought, so I tightened everything up to keep the settings and took one last image - but it would not solve. In the morning  I uploaded the image manually and it would still not solve until I processed it to make it clearer, must have been some thin cloud that I'd not noticed as the image timings were constant and all earlier attempts were solved fine.

So quite pleased the final settings, after tightening everything, ended up the best.

BTW.

A couple of points I thought worth mentioning:

• If a poor image is uploaded and the server takes too long while attempting to resolve should PPA handle that?
• Instead of having an extra button for Nova & Local for each image would it be tidier to have a select radio button to set either Local or Nova and have just one Solve for each?
• Not sure if using AstroArt5 as a solver will work, although it does have SDK for scripting I'm still having issues using it directly so would not hold out any hope of linking to it.  A local nova on the other hand sounds a good idea.

[themos]

Hello Stevie,

I've been thinking along the same lines. In my view, if you think a particular solve is taking too long, I would just abort the whole program and have a look at the image. There's very little chance of losing any work done as any previously solved files show up with a green "Solved". You gradually get a feel of what "too long" is. Local solves on my machine take about 6 seconds, for example. I am planning on having two configurations for local and adding a radio button to switch between the two (one for "do the most general thing", one for "here's a bunch of hints, I'm in a hurry"). I think AstroArt will work but you have to do a bit of work: save a JPEG file (as "img_01.JPG", say) and then rename the solved FITS file as "img_01.wcs". If this works well for people, I could finetune the process and have people just point to the AstroArt solved FITS file. 

Themos

[AndyKeogh]

I had a good session on the 18th Dec with almost 7 hours of good subs.

That was using v1.02 on Vista 32

I have brought the files over to my Win 8.1 desktop and tried running v1.03, but a previously solved file comes up as not solved It solves OK with a resubmission to Astrometry.net.

Checked back on my Vista laptop and it shows the original file as solved. So it appears that there is an issue with 1.02 solves running on 1.03.

Would you like me to mail you the image file and the 2 version of .wcs file?

Andy

[AndyKeogh]

I just tried again with the images that had been solved on V1.03 and the portrait image did not come up as solved. I still went ahead and asked it to find the polar axis which it did.

Looks like there is a glitch with the code that handles turning the solved from red to green.

Andy

[themos]

Hello Andy, I suspect the issue is the logic I use to make sure that the .wcs files are newer than the images.  When you copy files, the timestamps the operating system records (and that PPA sees) get changed. You can see the timestamps if you right-click on files and select Properties, select Details tab and scroll down to "Date created" and "Date modified" times. Could you tell me these for the  2 files on the source machine and the destination machine please?

[AndyKeogh]

Hello Andy, I suspect the issue is the logic I use to make sure that the .wcs files are newer than the images.  When you copy files, the timestamps the operating system records (and that PPA sees) get changed. You can see the timestamps if you right-click on files and select Properties, select Details tab and scroll down to "Date created" and "Date modified" times. Could you tell me these for the  2 files on the source machine and the destination machine please?

Hi Themos

I suspect you are in the right area but when I conducted the repeat test I was using files which had been solved on this computer. it also failed.

I rebooted my computer since I had removed python 2.7 32 bit. It is unfortunate that I cannot currently run PHD Lab and PPA as the former appears to require Python 2.7 in a specific directory

Still to support your investigation here is an image of the date created and date modified for each of my PPA files.

Despite this issue you have a cracking tool as I no longer worry about polar alignment but what causes the 16 pixel drift in my images over 7 Hours

Regards Andy

[AndyKeogh]

Looking at the last post, the first image showed the improvement though you cannot tell from the screenshot; maybe an area for improvement from a defensive programming point of view

Andy

[themos]

Andy, I have a separate Python for PHDLabs (in C:\Python2.7). There may be a way of using that one, once all the needed packages are added, but it's not straightforward.

On the issue of timestamps, I am now trying out this definition 

    def happy_with(self, wcs, img):        '''        check that .wcs (wcs) is compatible with .jpg (img)        '''        import os        from os.path import exists        if exists(wcs):            if os.stat(wcs).st_mtime> os.stat(img).st_mtime:                return True        return False

Perhaps you can try it in place of the existing function.

[Gonzo]

Wow Themos, nice work. How did I miss this thread???

I will definitely test this.

[Gonzo]

I've made a slight modification to the code (nothing special) to be able to run it from the command line under Linux/MacOsx :

macbook-air:PolarAlign stephanelucas$ head -n2 PPA.py#!/opt/local/bin/python macbook-air:PolarAlign stephanelucas$

This is to avoid having to type "python PPA.py" , I can now run it as ./PPA.py or even better move all the files to /usr/local/bin for example.

However, it seems that the file PPA.py was written on a Windows machine, had to dos2unix the file to remove non Unix characters.

If you are going to do the same, please check where your python is with the command "which python"

[themos]

Is there anything like Inno Setup for OSX? A free, easy-to-use tool for installing OSX apps? I see  https://developer.apple.com/library/mac/documentation/Darwin/Reference/ManPages/man8/installer.8.html is available always. If you know enough to hack an installation script, that would be great.

[D4N]

Interesting Gonzo, does this work in 10.10?

I made a script that I made executable so I can run the program from an icon.

cd "`dirname "$0"`"python PPA.py

TSED70Q, iOptron Smart EQ pro, ASI-120MM, Finepix S5 pro.

[JamesF]

If you want to use the OSX installer then you have to create a .pkg file, which is possible, but as far as I recall requires signing keys and registration with Apple etc.

What I (and many others) do is to package everything as a .dmg disk image which means it's just a "drag and drop" job to install where you want it.  However, that does mean that everything needs to be completely relocatable (because it can be installed anywhere) and it's "the Apple way" that all non system software dependencies are bundled with the package too.  That includes any external libraries that aren't part of the core OS installation and suchlike (which also may need paths changing internally).

That really seems to be what Mac users expect.  I get the impression that in the main they're not happy about being asked to run scripts or do anything from the command line.

James

[JamesF]

If you think it would be useful though, there's a build-dmg.sh script that comes with the sources for oacCapture.  By all means take that, rip out what might be useful to you and use it however you wish.  It's often much easier when you can see how someone else has done it to start with.

James

[AndyKeogh]

Andy, I have a separate Python for PHDLabs (in C:\Python2.7). There may be a way of using that one, once all the needed packages are added, but it's not straightforward.

When I had Python installed in C:\Python2.7 PHDLab worked but PPA did not -- complained of numpty being missing

On the issue of timestamps, I am now trying out this definition 

    def happy_with(self, wcs, img):        '''        check that .wcs (wcs) is compatible with .jpg (img)        '''        import os        from os.path import exists        if exists(wcs):            if os.stat(wcs).st_mtime> os.stat(img).st_mtime:                return True        return False

Perhaps you can try it in place of the existing function.

Well that worked a treat It has allowed me to complete my analysis of how stable polar alignment is over a 12 hour run. I also tried an east portrait and a west portrait to see if there was a difference.

Finally I tried a 30sec image "improvement" image instead of the normal 5 seconds but there was no significant difference in the reported error.

So position at the start of the imaging session was -- Polar error 0.58 arc min Left .52 Down 0.26

12 hours later again using an :-

East facing portrait -- Polar error 1.06 arc min Left 1.02 Up 0.28

Using a West Portrait -- Polar error 1.12 arc min Left 1.04 Up 0.40

Checked for an improvement with a 30 sec image -- Polar error 1.13 arc min Left 1.05 Up 0.40

So it looks as though it may pay me to align, move the scope around the sky for a bit and then recheck polar alignment again.

Andy

[Gonzo]

Interesting Gonzo, does this work in 10.10?

I made a script that I made executable so I can run the program from an icon.

cd "`dirname "$0"`"python PPA.py

TSED70Q, iOptron Smart EQ pro, ASI-120MM, Finepix S5 pro.

It does work in 10.10 yes.

[Gonzo]

Is there anything like Inno Setup for OSX? A free, easy-to-use tool for installing OSX apps? I see  https://developer.apple.com/library/mac/documentation/Darwin/Reference/ManPages/man8/installer.8.html is available always. If you know enough to hack an installation script, that would be great.

I have no idea how to do it, will Google it and see if it can be done easily.

[Gonzo]

Found that with py2app on Mac you can create a standalone app out of your Python scripts. Found the following interesting tutorial.

Edited December 22, 2014 by Gonzo